Applying Machine Learning to Economic Analysis: Predicting Bank Failures in India Using Decision Tree Models

This study examines the application of machine learning (ML) techniques, specifically decision trees (DTs), to predict the financial stability of institutions. Using the C5.0 algorithm, a DT was constructed and tested on a dataset, achieving an accuracy of 89.7%. Key financial indicators, including financial margin and profitability of assets, were identified as the most influential in determining success or failure. The findings highlight the effectiveness of DTs in economic forecasting, demonstrating their potential for financial risk assessment. However, limitations such as data availability, the lack of time-series analysis, and the model's inability to learn incrementally were identified. Despite these challenges, the results suggest that DTs can serve as a valuable tool for financial decision-making. Future research can refine the model by incorporating additional economic variables and exploring alternative ML algorithms to improve predictive accuracy.